Answers are grounded in this paper and cite their sources.
Numerous studies emphasise the importance of adopting an intersectional perspective to examine unwarranted disparities in the criminal justice system. However, traditional methods, such as interaction effects, often result in overfitted models. Consequently, most studies focus on disparities across broad groups of offenders, neglecting the wider range of unique experiences that result from the intersection of multiple socio-demographic characteristics. Here, we use a new modelling approach, the multilevel analysis of individual heterogeneity and discriminant accuracy (MAIHDA), to explore the presence of intersectional disparities in sentencing that arise from the combination of offenders' gender, race, level of education, and citizenship. Our findings indicate that these socio-demographic factors alone explain up to 19% of the variability in imprisonment decisions and 14% of the variability in sentence length within US Federal Courts. Additionally, we find that standard models ignoring intersectional disparities fail to account for up to 20% of the unwarranted disparities detected. These findings highlight the importance of placing intersectionality at the centre of the study of sentencing disparities and demonstrate the potential of MAIHDA to do so more comprehensively than previously possible.
No takes yet. Share an insight, caveat, or question.
Pina-Sánchez et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: